In addition to the issue with multiple growth morphs, I would like to call to your attention other changes in 3.30 and the converter that could affect some of you.
One is the removal of the lower limit of 1.0 on the sample size for composition data. So, if your approach is to apply iterative-re-weighting, some of your composition observations could get down-weighted below a sample size of 1.0.
The other is with regard to the distribution of recruitment among areas and time. The issue was that 3.24 conceived of recruitment distribution by "birth seasons" so it needed a parameter for each season in a multi-season model. In 3.30, that concept is replaced by settlement event and the number and timing of settlement events is now much more flexible and it changes the number of expected parameters. In getting this clarified, I used it as an opportunity to get rid of the notion that option 1 (birthseasons) still existed in 3.30. The converter makes the change for you. Also be sure to turn off the recruitment distribution interaction flag. It should be 0. If you have a complex situation seeming to need such an interaction, the better option is now #3 that has a simple list aligning one parameter with each settlement event. So the simplest setup in 3.30 now looks like:
2 # recr_dist_method for parameters: 2=main effects for GP, Area, Settle timing; 3=each Settle entity
1 # not yet implemented; Future usage: Spawner-Recruitment: 1=global; 2=by area
1 # number of recruitment settlement assignments
0 # unused option (formally in 3.24 this was the recruitment distribution interaction flag)
#GPattern month area age (for each settlement assignment)
1 1 1 0
The third issue is the addition of "age" to the set of inputs on the above line. The issue is that SS3.24 had all fish recruiting at an integer age of 0. Unfortunately, in a multi-season setup with spawning in season 2 and recruitment the following year in season 1, the recruits still started at age 0, which may or may not be logical. Now in 3.30 we have spawning specified by month, say month 9. Then if we have some of the recruits settling in December (month 12) they will start as age 0 and progress to age 1 on Jan 1 (even though only 1 month old). If another settlement event is set to happen on January 15 (month 1.5), then we want these fish to start at age 1 so they will be correctly placed in the numbers-at-age matrix to align with the same cohort that settled a month earlier in December. These recruitment settlement enhancements are a new capability of SS3.30 and I look forward to seeing how you can use them. I would like to see a project investigate how a spread in settlement timing events can create size-at-age data that mimic what we currently perceive as higher CV of size-at-age for young fish.
#GPattern month area age (for each settlement assignment)
1 12.0 1 0
1 1.5 1 1
